Usually, "piloting error" is the preferred cause by aircraft manufacturers and bean counters so you have to read between the lines unless the investigating authority is an honest one.
What strikes me here is that the airplane only climbs approx. by 2500' during G/A before starting to descend towards impact. I think that a readout from the FDR will reveal the cause pretty quickly. Does anyone have the published G/A procedure for the runway?